A Multilevel Analysis of the Determinants of Under-Five Mortality in Nigeria: Individual, Household, and Community-Level Factors

Under-five child mortality in Nigeria remains the highest in absolute terms globally, and its determinants operate simultaneously at individual, household, and community levels in ways that single-level analytical approaches inadequately capture. This study conducted a multilevel analysis of determinants of under-five mortality, integrating individual, household, and community-level factors using nationally representative data. A secondary data analytical study was conducted using merged 2023 NDHS individual recode, household recode, and geographic cluster-level datasets for a birth history cohort of 18,420 children born in the five years preceding the survey. Cox proportional hazards multilevel frailty models were constructed to account for within-cluster correlation. Community-level predictors were derived from cluster-level aggregation of key determinants and satellite-derived poverty and infrastructure data. Individual-level predictors included birth size, breastfeeding status, and vaccination status. Household-level predictors included wealth quintile, maternal education, and water and sanitation access. Community-level predictors included cluster-level vaccination coverage, healthcare facility density, rural-urban classification, and poverty index. Results showed that community-level factors explained 21.3 percent of total mortality variance. Community vaccination coverage showed the strongest community-level protective effect (HR 0.54 per 10 percentage point increase; p < 0.001). Household wealth and maternal education retained significant independent effects at all levels after mutual adjustment. Significant cross-level interactions between maternal education and community healthcare density were identified, showing stronger maternal education effects in communities with denser healthcare facilities. The study provides the most comprehensive multilevel under-five mortality analysis using 2023 NDHS data and recommends community-targeted child survival strategies. Keywords: under-five mortality, multilevel analysis, determinants, Nigeria, child survival